Multivariate approach to QRS detection
We have developed a QRS detection algorithm based on a multivariate model, in which three independent, normalized measures -- amplitude, first difference, and spatial frequency -- are combined in a weighted sum to generate an indicator variable that is then compared to a detection threshold. To increase sensitivity, we first applied and FIR, band-pass filter consisting of a cascaded series of running medians and means. The techniques appears to be exceedingly robust, correctly detecting even aberrant QRS complexes in noise-corrupted ECGs.
Proceedings of the Annual Conference on Engineering in Medicine and Biology
Bond, A. B., Greco, E. C., Bowser, R., Kadri, N. N., & Sketch, M. H. (1993). Multivariate approach to QRS detection. In Proceedings of the Annual Conference on Engineering in Medicine and Biology 15 (pt. 2): 675-676.